Page 127 - Becoming Metric Wise

118 Becoming Metric-Wise

this is combined with a low self-cited rate this may suggest that the
researcher is a productive, key figure in his research specialty.
A journal’s self-citing and self-cited rates can be obtained from data
available in the JCR.
Porter (1977) found that it does not matter whether one includes
author self-citations or not, at least when using citation analysis to study
science subfields or any other large unit. Gla ¨nzel et al. (2004) point out
that author self-citations are an organic part of the citation network.
They can be measured and modelled and consequently, significant devia-
tions of reference standards can be determined.

5.6.3 Coauthorship and Counting Contributions

We will now turn to the difficult problem of coauthorship and its influ-
ence on counting contributions (of authors, institutions, countries).
Usually, scientists consider each article with their name in the by-line as
one of their articles. For an outsider, however, this is not so straightfor-
ward. The productivity of author A who wrote four articles on her own,
is certainly larger than that of author B who wrote four articles in collab-
oration with three colleagues (during the same period).
Generally speaking there are nowadays five major approaches for
counting contributions depending on the number and/or position of
authors in the by-line:
• First author count (also called straight count) in which only the first
author receives a credit.
• Major contribution count in which the first author(s) and the corre-
sponding author(s) (if different) receive a credit. One either gives a
full credit (one) to each major contributor; or credits are fractionalized
over major contributors.
• Complete or normal count, giving each contributing author a full
credit.
• Complete-normalized fractional counting in which each coauthor
receives 1/n credits for an article with n authors.
• Other methods for fractional counting are proposed in the literature,
among which harmonic counting is the most interesting one, see
further.
Counting methods can be subdivided into four groups, according to
two times two criteria. The first makes a distinction between giving
credit to each author and sometimes giving no credit to some authors.